JP5062634B2 - Base sequence design method - Google Patents

Description

  The present invention relates to a method for designing a base sequence of a gene for expression in a translation system.

  When a gene encoding a specific amino acid sequence is expressed in a host cell of a species different from the species from which the gene is derived, the translation process becomes a rate-limiting step, and a sufficient expression level may not be obtained.

  As one of the solutions in such a case, the codon used by the gene has been changed to a codon frequently used in the host cell species.

  However, there are cases where the expression level does not become sufficiently high, and further ingenuity is expected in the design of the gene to be expressed.

  Therefore, an object of the present invention is to provide a gene base sequence design method capable of obtaining a high expression level when a gene is expressed using a predetermined translation system.

  When a gene is expressed using a translation system, the gene is often directly incorporated into an expression vector and expressed in the translation system regardless of the species from which the gene and translation system are derived. However, since codon usage differs depending on the species, the codon of the gene may not be used well in the translation system used, and the expression level may not increase. As a result of diligent efforts to solve this problem, the present inventors have found a new finding that the regularity of the codon sequence varies depending on the species, and this is one of the reasons that the expression level does not increase. The present invention has been completed.

  The nucleotide sequence design method of the present invention is a nucleotide sequence design method for a gene to be expressed using a predetermined translation system, characterized in that the codon sequence of the gene is adapted to the regularity of the codon sequence in the learning data. To do. The learning data may be a gene pool of a species from which the translation system is derived, or may be a gene pool whose expression level has already been clarified to be a predetermined level or higher in a predetermined translation system. Good.

  The base sequence design method may include a step of adding the adapted base sequence of the gene as data to the learning data used for the adaptation.

  Furthermore, the base sequence design method may be characterized in that the codon sequence of the gene is adapted to the regularity of the codon sequence in the learning data using a hidden Markov model. In this case, the regularity of the codon sequence may be calculated using a Baum-Welch algorithm.

  Furthermore, in the base sequence design method, the translation system may be in a cell or in an in vitro protein synthesis system.

  The program of the present invention is a program for designing a base sequence of a gene to be expressed using a predetermined translation system, and is a program for causing a computer to execute the base sequence design method. The recording medium of the present invention is a computer-readable recording medium that records at least one of these programs.

  The base sequence design system of the present invention is a gene base sequence design system for expression using a predetermined translation system, comprising: (i) an input device for inputting base sequence data; and (ii) input A computer that executes the program according to claim 9 by using the data, and an output device for outputting the result obtained by (iii) and (ii) are provided.

== Cross reference with related literature ==
Note that this application claims the benefit of priority based on Japanese Patent Application No. 2006-64816 filed on Mar. 9, 2006, and is incorporated herein by reference.

It is a schematic diagram which shows the specific example of three conditions which are examples of the "pruning" in embodiment concerning this invention.

  Unless otherwise stated in the embodiments and examples, J. Sambrook, EF Fritsch & T. Maniatis (Ed.), Molecular cloning, a laboratory manual (3rd edition), Cold Spring Harbor Press, Cold Spring Harbor, New York (2001); FM Ausubel, R. Brent, RE Kingston, DD Moore, JG Seidman, JA Smith, K. Struhl (Ed.), Standard Protocols in Molecular Biology, John Wiley & Sons Ltd. The method described in the protocol collection, or a modified or modified method thereof is used. In addition, when using commercially available reagent kits and measuring devices, unless otherwise explained, protocols attached to them are used.

  The objects, features, advantages, and ideas of the present invention will be apparent to those skilled in the art from the description of the present specification, and those skilled in the art can easily reproduce the present invention from the description of the present specification. it can. The embodiments and specific examples of the invention described below show preferred embodiments of the present invention and are shown for illustration or explanation, and the present invention is not limited to them. It is not limited. It will be apparent to those skilled in the art that various modifications and variations can be made based on the description of the present specification within the spirit and scope of the present invention disclosed herein.

== Base sequence design method ==
The present invention is a method for designing a base sequence of a gene to be expressed using a predetermined translation system, characterized in that the codon sequence of the gene is adapted to the regularity of the codon sequence in the learning data. This will be specifically described below. In the present specification, “codon sequence” means “a sequence of triplet codons”. Further, “regularity of codon sequences” means “codon sequences with the highest probability of being expected to appear”.

  When expressing a target gene using a predetermined translation system, the gene is directly incorporated into an expression vector that can be expressed in the translation system regardless of the species from which the translation system is derived or the species from which the gene is derived. It is often introduced in. However, according to the present invention, by adapting the codon sequence of the gene to the regularity of the codon sequence in the learning data, the base sequence of the gene can be optimized for expression in the translation system, and the expression level can be increased. Can be raised.

  The translation system to be used is not limited as long as it is a system capable of translating mRNA, and may be in a cell or an in vitro translation system. When using an intracellular translation system, for example, a gene whose sequence is optimized may be incorporated into an expression vector and introduced into the cell by a conventional method. When using an in vitro translation system, for example, mRNA synthesized by transcription in vitro may be added to a rabbit reticulocyte, wheat germ, or cell-free protein synthesis system derived from E. coli and translated.

  The gene to be expressed is not particularly limited, and any amino acid sequence may be used, and it may be a natural gene or an artificially introduced gene.

  The learning data is not particularly limited, but there is a gene pool of a biological species or a related species from which the translation system is derived, or a gene pool whose expression level has already been clarified in the predetermined translation system. Preferably, they may be combined. In addition, when the genetic data in which the codon sequence is adapted by the base sequence optimization method of the present invention is added to the learning data used in the adaptation, and the same translation system is used next, the learning data to which the data is added is added. It may be used.

  The base sequence is optimized by adapting the codon sequence of the gene to be expressed to the regularity of the codon sequence in the learning data.

  For example, consider the case where the codon sequence of a gene to be expressed is determined by the regularity of two codon sequences. As an example, a sequence of two amino acids in a gene to be expressed is A-B. If A has three types of codons a1, a2, and a3, and B has two types of codons, b1 and b2, there are six ways to select codons. In learning data, if the frequency of a1-b2 is the highest, a1-b2 is selected as the base sequence of the gene to be expressed. As described above, in the learning data, the codon sequence having the highest probability of being expected to be selected is selected as the codon sequence of the gene to be expressed.

  In this example, the number of codons in the codon sequence that considers regularity is two, but it may be three or more, and considering the compatibility between the learning data and the codon sequence of the gene to be expressed, we think that the larger number of codons is better. It is done. When the hidden Markov model is used, the conformity of a codon sequence having an arbitrary length can be calculated theoretically, but the number of codons in the codon sequence considering regularity is determined by the number of hidden states. As the number of codons increases, the amount of calculation of the frequency of codon sequences increases and the cost also increases. Accordingly, the number is preferably 10 to 50.

A method for defining the regularity of the codon sequence in the learning data is not particularly limited, but it is preferable to use a hidden Markov model. In the hidden Markov model, a large amount of training data {g ¹ , ..., g, ..., g ^T } is collected, and for codons c _i in all these genes,
Find _a set of parameters {π _a , s _ab , d _a (c _i )} that maximizes. (Where π _a is the initial probability of state a before output of the start codon (Met), s _ab is the state transition probability from state a to state b, and d _a (c _i ) is state a and codon c _i is observed. Output probability.)

Here, each parameter is preferably determined using a Baum Welch algorithm. That is,
1. Set initial values π ₁ , s _ab , d _a (c _i ).
2. In the gene g on the learning data, the transition probability γ _i (a, b, g) from the state a to the state b at the i-th codon is calculated from the learning data for all combinations of i, a, b.
Specifically, the following equation is calculated (where L is the total number of codons for gene g, c _i ^L is a character string meaning the i-th to L-th (that is, the last) codon, and S is the state a. corresponding to all the state transitions with respect to s _ab corresponding to the state transition to state b from s _ab, D codon c _i d _a that corresponds to the (c _i) d corresponding to all codons with respect to _a (c _i) .)
3. Calculate and redefine π ₁ , s _ab , d _a (c _i ) from γ _i (a, b, g).
Specifically, the following formula is calculated.
4. Return to 2 if each parameter is not stable.

  This algorithm may be performed until each parameter is completely stabilized. However, since it takes an enormous amount of time, this algorithm may be terminated a predetermined number of times. The number of times in this case may be about 100 to 1000 times. In 1, the initial value setting method is not particularly limited, but may be set by adding a probabilistic constraint to the pseudo-random number as in the embodiment.

Using each parameter obtained by the Baum Welch algorithm, calculating the output probabilities for all the codon combinations, and obtaining the maximum one, the target codon sequence can be obtained. That is, the output probability of the character string in the model M determined by the Baum Welch algorithm is the sum of all combinations of the state transitions σ _{1, a} , σ _{2, a2 ,.}
It becomes.

However, since this calculation takes an enormous amount of time, a method for obtaining an approximate value may be used, and the method is not limited. For example, “pruning” having the following three conditions is used. May be. (Note that FIG. 1 shows specific examples of these three conditions.)
First, it is clear about (i), and since the purpose here is to determine the codon sequence of the gene to be expressed, the amino acid sequence to be targeted is determined in advance, and the i-th amino acid a _i Although it can be said that (ii) cannot select a base sequence that uses a codon c _i that does not correspond to, it will be proved strictly by induction below.

  As described above, (i) and (ii) do not make an error in obtaining a suitable solution, but (iii) can make an error in obtaining a suitable solution. By sufficiently increasing n, the condition of (iii) becomes strict and the possibility of erroneous pruning is reduced. Therefore, it is preferable to increase n sufficiently. However, as n is increased, the amount of calculation required to perform the check (ii) increases.

== Base sequence design system ==
In order to automate the above base sequence design method, it is programmed so that it can be executed by a computer. The program created in this way is also within the scope of the rights of the present invention.

  Furthermore, a base sequence design system including an input device for inputting base sequence data and an output device for outputting the result obtained by executing the program together with a computer for executing the program may be used. it can.

<Example>
(1) Design of nucleotide sequences of adh1 gene and adh2 gene of S. cerevisiae In this example, Escherichia coli JM109 was used as a translation system. Since this cell line is derived from Escherichia coli K12, the base sequence data of E.coli K12 MG1655 in the genome database DDBJ (http://www.ddbj.nig.ac.jp/) is used as learning data. used. As genes to be expressed in this translation system, alcohol dehydrogenase genes adh1 (adh1w, SEQ ID NO: 1) and adh2 (adh2w, SEQ ID NO: 2) derived from Saccharomyces cerevisiae were used.

  First, the number of codons in the codon sequence considering regularity was set to 8, and the regularity of the codon sequence up to 8 amino acids before the amino acid determining the codon was examined. In order to define the regularity of the codon sequence in the learning data, a hidden Markov model was used. In the Baum Welch algorithm, the initial value is a random number that is set so that the sum of all probability values of the state transition probability and output probability is 1, compared to the pseudo-random number generated by inputting the time. Was used. In addition, the learning data was learned by performing an algorithm of 350 times.

 Using the model thus obtained, the codon sequences for Adh1 and Adh2 were optimized. At that time, “pruning” having the above three conditions was used. In this example, when n was set to 50, the calculation of (iii) did not occur in both Adh1 and Adh2 derived from S. cerevisiae. Therefore, it is not necessary to perform pruning that may contain errors. The exact optimal solution was obtained.

(2) Synthesis of optimized adh1 gene and adh2 gene The optimized adh1 gene (adh1c, SEQ ID NO: 3) and adh2 gene (adh2c, SEQ ID NO: 4) were synthesized as follows.

  For adh1c, synthesis was performed by a method of joining 40 bp primer groups (Assembly PCR method). First, a forward primer and a reverse primer corresponding to the base sequence of adh1c were prepared by cutting them into 40 bp lengths. Since the length of the base sequence of adh1 was 1047 bp, it became 26 forward primers (SEQ ID NOs: 5 to 30) and 26 reverse primers (SEQ ID NOs: 31 to 56). The 26 × 2 primers prepared in this way were subjected to assembly PCR under the following conditions.

First, a primer mix (250 μM) in which all the primers were mixed at an equal concentration was prepared, and 55 cycles of 94 ° C. for 30 seconds−52 ° C. for 30 seconds−72 ° C. for 30 seconds were performed with the following reaction solution (50 μl).
Reaction solution: 0.2 mM dNTP 5 μl
primer mix 0.5μl
Pfu Ultra (Stratagene) 0.25μl
10 × buffer 5μl for Pfu Ultra
Sterile water 39.25μl

After completion of the reaction, PCR was performed again with the following reaction solution using the reaction solution as a template. The conditions were 94 ° C. for 2 minutes denaturation reaction, 94 ° C. for 30 seconds-53 ° C. for 30 seconds-72 ° C. for 60 seconds for 30 cycles, and 72 ° C. for 10 minutes for extension reaction. At this time, a primer (SEQ ID NO: 57) whose length was adjusted by attaching an EcoRI site at the beginning of the first forward primer (SEQ ID NO: 5) so that ligation with the vector can be performed at a later stage as a primer. A reverse primer (SEQ ID NO: 31; EcoRI site added) was used.
Reaction solution: Template 1μl
0.2 mM dNTP 5 μl
Primer (SEQ ID NO: 57) (250 μM) 0.5 μl
Primer (SEQ ID NO: 31) (250 μM) 0.5 μl
Pfu Ultra (Stratagene) 1μl
10 × buffer 5μl for Pfu Ultra
Sterile water 36μl
The PCR product was purified on an agarose gel, cut with EcoRI, and inserted into the EcoRI site of pUC19. By determining the base sequence, it was confirmed that it had the correct sequence. Regarding adh2c, the synthesis of a plasmid designed in the same manner as adh1c was outsourced to operon, but basically it can be synthesized in the same manner as adh1c.

(3) Synthesis of adh1 gene and adh2 gene On the other hand, adh1w and adh2w were subjected to PCR using genomic DNA extracted from S. cerevisiae YPH 499 as a template to obtain each clone. PCR uses a primer pair for each gene (a primer pair of SEQ ID NOS: 58 and 59 for adh1w, a primer pair of SEQ ID NOS: 60 and 61 for adh2w), and denaturation at 94 ° C. for 2 minutes. After the reaction, 30 cycles of 94 ° C. for 30 seconds-53 ° C. for 30 seconds-72 ° C. for 60 seconds were performed, and an extension reaction was performed at 72 ° C. for 10 minutes.
Primer (capital letters indicate restriction sites added for cloning):
adh1w amplification forward primer (SEQ ID NO: 58):
GGAATTCatgtctatcccagaaactcaaaaaggtgtt
adh1w amplification reverse primer (SEQ ID NO: 59):
GGAATTCttatttagaagtgtcaacaacgtatctacc
adh2w amplification forward primer (SEQ ID NO: 60):
GGAATTCatgtctattccagaaactcaaaaagccatt
adh2w amplification reverse primer (SEQ ID NO: 61):
GGAATTCttatttagaagtgtcaacaacgtatctacc
Reaction solution: Template 1μl
0.2 mM dNTP 5 μl
Forward primer (250 μM) 0.5 μl
Reverse primer (250 μM) 0.5 μl
Pfu Ultra (Stratagene) 1μl
10 × buffer 5μl for Pfu Ultra
Sterile water 36μl
The PCR product was purified on an agarose gel, cut with EcoRI, and inserted into the EcoRI site of pUC19. By determining the base sequence, it was confirmed that it had the correct sequence.

(4) Construction of expression vectors for each gene of adh1w, adh2w, adh1c, and adh2c PCR was performed using the plasmid obtained by inserting adh1c, adh2c, adh1w, and adh2w thus obtained in the EcoRI site as a template. PCR was carried out under the same conditions as in (3), except that the primer was a primer pair that specifically amplifies each gene (for adh1c, the primer pair of SEQ ID NOS: 62 and 63, and for adh2c, the sequence The primer pair of numbers 64 and 65, the primer pair of SEQ ID NOs: 66 and 67 for adh1w, and the primer pair of SEQ ID NOs: 68 and 69 for adh2w) were used.
Primer (capital letters indicate restriction sites added for cloning):
forwaord primer for adh1c amplification for pCold (SEQ ID NO: 62)
CCGGAATTCatgagcattccggaaacgcaaaaaggtgtg
A reverse primer for adh1c amplification for pCold (SEQ ID NO: 63)
AACTGCAGttatttgctggtatccaccacatagcggcc
forwaord primer for adh2c amplification for pCold (SEQ ID NO: 64)
CCGGAATTCatgagcattccggaaacccagaaagcgattatt
Reverse primer for adh2c amplification for pCold (SEQ ID NO: 65)
AACTGCAGttatttgctggtatccaccacatagcggcc
forwaord primer for adh1w amplification for pCold (SEQ ID NO: 66)
CCGGAATTCatgtctatcccagaaactcaaaaaggtgttatc
reverse primer for adh1w amplification for pCold (SEQ ID NO: 67)
AACTGCAGttatttagaagtgtcaacaacgtatctacc
forwaord primer for adh2w amplification for pCold (SEQ ID NO: 68)
CCGGAATTCatgtctattccagaaactcaaaaagccattatc
reverse primer for adh2w amplification for pCold (SEQ ID NO: 69)
AACTGCAGttatttagaagtgtcaacaacgtatctacc
The PCR product was purified on an agarose gel, cleaved with EcoRI and PstI, and then inserted into the EcoRI-PstI sites of pColdIII and pColdIV (Takara Bio), which are low-temperature expression-inducing vectors. The plasmid thus prepared (Table 1) was transformed into JM109.

(5) Expression induction of the optimized adh1 and adh2 genes First, the E. coli JM109 strain having each plasmid and the JM109 strain having pUC19 without an insert in 3 ml LB liquid medium for about 12 hours at 37 ° C Pre-cultured with stirring. Subsequent main culture operations differ depending on the vector as follows.

  For the plasmid containing pUC19 as a vector, after pre-culture, add all 3 ml of the culture solution to 100 ml LB liquid medium (+ ampicillin 50 μg / ml) in a Sakaguchi flask and shake culture at 37 ° C for 1.5 hours (120 rpm), IPTG was added to a final concentration of 100 μM, and then the shaking culture (120 rpm) was performed again at 37 ° C. for 24 hours.

  For the plasmid containing pCold as a vector, after pre-culture, add 3 ml of all culture solution to 100 ml LB liquid medium (+ ampicillin 50 μg / ml) in a Sakaguchi flask and shake culture at 37 ° C for 1.75 hours (120 rpm) and then allowed to stand at 15 ° C. for 30 min. Thereafter, IPTG was added to a final concentration of 100 μM, and the mixture was further cultured with shaking (120 rpm) at 15 ° C. for 24 hours.

 The culture solution thus obtained was centrifuged at 10,000 rpm for 10 min, and the cells remaining after discarding the supernatant were suspended in 100 mM Tris buffer (pH 8). The suspension was centrifuged again at 7,200 rpm for 10 min, and the cells remaining after discarding the supernatant were suspended in 100 mM Tris buffer (pH 8). The suspension thus obtained was subjected to ultrasonic crushing at 4 ° C. for 10 min, and then centrifuged again at 10,000 rpm for 10 min, and the supernatant was used as a cell-free extract.

The ADH activity in this cell-free extract was measured as follows. That is, the cell-free extract was diluted 1-fold, 10-fold, 100-fold and 1000-fold, 100 μl of each was mixed with 900 μl of the following substrate solution, and the absorbance at a wavelength of 340 nm was measured up to 50 sec at 10 sec intervals.
Substrate solution: Sodium pyrophosphate 85 mM
Semicarbazide hydrochloride 6.5 mM
Glycine 18mM
Ethanol 550mM
NAD ⁺ 1.75mM

For each cell-free extract, the total protein amount was measured using the Bradford method (CBB method) with BSA as a standard. ADH activity was calculated from the obtained absorbance using the following formula. Table 2 shows the results. The specific activity of the negative control (E. coli JM109 strain without any plasmid and E. coli JM109 strain with pUC19 without insert) was 0.

  Thus, by optimizing the base sequence, the expression level of the protein was enhanced 2.5 to 10 times except for one combination (adh2 and pColdIV). From the above, it was shown that when a gene is expressed using a predetermined translation system, the expression level can be enhanced by at least 2.5 to 10 times by the base sequence design method of the present invention.

  According to the present invention, it is possible to provide a gene base sequence design method capable of obtaining a high expression level when a gene is expressed using a predetermined translation system.

Claims (9)

A method for designing a base sequence of a gene to be expressed using a predetermined translation system,
An input device inputting a gene base sequence into a computer;
The computer, the codon sequences obtained from the nucleotide sequence of the gene that has been input, the training data, the regularity of the calculated codon sequence using Baum-Welch algorithm, Ru fitted using Hidden Markov Models step And
An output device outputting the adapted base sequence of the gene;
Nucleotide sequence design method, which comprises a.   The method according to claim 1, wherein the learning data is a gene pool of a biological species from which the translation system is derived.   The method according to claim 1, wherein the learning data is a gene pool whose expression level has already been clarified to be equal to or higher than a predetermined level in a predetermined translation system. Said computer further, the base sequence of the gene is adapted to the learning data used in adapting method of claim 1, characterized in that it comprises a step of adding as data. The method according to claim 1, wherein the translation system is in a cell. The method according to claim 1, wherein the translation system is in an in vitro protein synthesis system. A program for designing a base sequence of a gene to be expressed using a predetermined translation system,
The program which makes a computer perform the method of any one of Claims 1-6 . A computer-readable recording medium on which the program according to claim 7 is recorded. A gene base sequence design system for expression using a predetermined translation system,
(i) an input device for inputting the base sequence of the gene into a computer ;
(ii) The codon sequence obtained from the base sequence of the gene input by the input device is adapted to the regularity of the codon sequence calculated using the Baum-Welch algorithm in the learning data using a hidden Markov model. A computer for, and
(iii) A base sequence design system comprising an output device for outputting the adapted base sequence of the gene .